Congratulations to both Hikaru Nakamura and Irina Krush are in order. They won the U.S. Championship and the U.S. Women's Championship, respectively, and pocketed some nice coin along the way as well - $45k for Nakamura and $20k for Krush.
Nakamura entered the last round half a point ahead of Ray Robson, and that's how he ended it too. Robson's game finished after Nakamura's, but that Robson would win was clear early on. He faced Timur Gareev, who has been playing somewhat eccentric chess throughout the tournament; today, he went completely out of orbit and left the solar system. Gareev is a great player whose FIDE rating reached a high of 2682 a couple of years ago, so his managing to achieve a lost position with white in just nine moves should be chalked up his having too much talent rather than too little. The game lasted 31 moves, but the final result was seldom if ever in doubt.
In the meantime, Alexander Onischuk managed to equalize against Nakamura pretty comfortably, and was probably just a few precise moves away from making a draw. Had he done so, Nakamura would have had a rapid playoff against Robson on Monday. Instead, Onischuk made a few little inaccuracies, got in trouble, and then was lost - all within a space of six or seven moves. The fatal moment came for Black when he played 27...Nxf2, which was a tactical blunder. Instead, 27...Rb6 would have led to a double rook ending where White's extra pawn would give him the ability to torture Black for many moves to come, but not probably not win against correct defense. After 27...Nxf2 28.Nd4 Nh3+ Onischuk probably missed Nakamura's 29.Kh1!, winning (at least) an exchange, and he resigned two moves later.
Nakamura thus finished in clear first with 8/11, Robson in clear second with 7.5 points, and in clear third was Wesley So with 6.5 after another strong win, this time against Kayden Troff. It's to So's credit that he finished so well, and it's to his relief (I hope) that the final margin was such that even if So had defeated Akobian he still would have finished half a point behind Nakamura.
In the remaining games, Shankland drew with Akobian, Naroditsky drew with Kamsky and, perplexingly, Sam Sevian beat Conrad Holt. It's difficult to be certain about this, in part because the live commentary ended shortly after Nakamura's win and most of what we have to go by is the very fallible result (probably) generated by the DGT boards. Here's the data I have: the tournament website's crosstable and the broadcast board on Chess24 both give the result as a win for Sevian. Moreover, the broadcast board shows Black (Holt) having no time, which offers a reasonable explanation. But look at the game itself: absolutely nothing happened from move 52 until move 99, when Holt finally decided to push his a-pawn up a square to avoid a coming 50-move rule claim, and a move later he lost on time in an absolutely safe position two pawns ahead. If they had been playing without an increment then sure, things like this can happen, but with 30 seconds added after every move Holt must have just lost track of the clock. This is possible, but it's also possible that he just decided that playing the position out was pointless and agreed to a draw, and they didn't bother to stop the clock afterward and/or the kings were put on the wrong squares in the center, at least momentarily.
We're in rich tangent territory here, and I'll indulge a bit before turning briefly to the women's championship. If it turns out that this is just another DGT error, it might be time to hire some unemployed people to picket that company's headquarters until they make some sort of design fix. But rather than beat that dead horse, here's a new topic: why did the English-language commentators, who were on site, quit their broadcast so early? I'm not sure if the culprit is the St. Louis club or Yasser Seirawan, but this happened all tournament long. I'm sure the Sinquefields are paying him well; is it too much to ask that he (and Jennifer Shahade & Maurice Ashley) actually stay for the entire time? It is frankly incredible to me that the Spanish-language commentary coming from a Chess24 studio in Europe more than once outlasted the on site English-language commentary team based at the tournament site itself. Even if Yaz & crew didn't stick around for all 100 moves of Sevian-Holt, they didn't have to pack their bags when just two of the six games in the open section had finished. The St. Louis club is doing some great things for chess in the United States, but there are some things they could do better - and this is one of them.
Turning back to the game itself, I should note one especially interesting moment, which would have preempted all of this discussion. Holt was better almost all the way, but a big slip on move 26 gave Sevian the chance to finish in style. 27.Nf5! would have won, threatening 28.Qh8+ followed by 29.Rh7#. The main variation runs 27...exf5 (creating a flight square on e6) 28.e6! (preventing the king from running, after 28...Qxe6 the flight square is gone) 28...Qg7 29.Kf2 (threatening Rh1 followed by Rh8+, mating) and wins. Black must play either 29...Nd8, when after 30.e7! he's going to lose practically everything (30...Qe7 31.Rh8+ followed by 32.Rh7+, 33.Rxe7(+) and 34.Qxb7), or he plays 29...Bf/d7 30.Rh1! Bxe6 31.Rh8+ Kf7 32.Qc7+! Ne7 33.R8h7, when one funny finale is 33...Rbe8 34.Qe5! with mate in three.
As for the strange way Holt lost - if he lost - it reminds me of one of my luckiest wins ever. After making a huge error in a game where I was clearly better, I had to go into a bad ending a pawn down against a strong expert (approximately 2140 USCF). There were many further adventures in the game, but I somehow reached an ending with king and rook against my opponent's king and queen. This is a theoretical win for the queen, of course, but finishing off the rook isn't trivial against good defense. (Walter Browne initially failed to defeat a computer in that ending in a specially arranged challenge, and years later Peter Svidler once famously failed to win this ending against Boris Gelfand in a crucial FIDE knockout world championship match, in a rapid playoff.) Unfortunately for my opponent, he had only seven seconds to win it, but with a five second time delay every move. (That is, there was a five second grace period each move before the seven seconds would start ticking off.) He made most of his moves without losing any of his time, but at one point he burned five of his "real" seconds, and then around 25 moves into the endgame he spent his last two seconds, and lost. It was a difficult situation for my opponent: just making moves would have been easy, but to make progress against good defense one must concentrate. And once one really concentrates, how does one remember to move?
So perhaps something like that happened to Holt. Another topic for discussion: should he have kept playing that ending against Sevian, after not making a shred of progress for 48 moves? Normally I'd say that he had carried out the appropriate desire to fight for a win a bit too far, but something important was at stake: a place in the World Cup this September. I'm not sure what the tiebreak situation was in case of a draw: Troff, Holt and Sevian would have all had 5 points and only one last spot was available. (Shankland also had 5 points, but had already qualified from another event.) If Holt had won, the spot would have been his; instead, it's the 14-year-old Sevian who has qualified, along with Nakamura and So (by rating), Shankland (from the American Continental Championship), and Robson, Onischuk, Akobian, and Kamsky from this event.
Time for a few words about the women's championship. Irina Krush led her closest rivals by a full point entering the last round, and needed only a draw with white against one of them - Katerina Nemcova - to seal the deal. This she managed to do in an efficient and expeditious manner, and she has now won her 7th U.S. Women's crown and fourth in a row. Very impressive! - but it's not the record. Gisela Kahn Gresser (1904-2000) won it nine times from 1944 to 1969; an especially impressive feat considering that she didn't even learn how to play until she was in her 30s! Krush, by contrast, only turned 31 in December, so the odds are very good she will exceed Gresser's mark and then some before she decides to call it a career.